[personal profile] tangaroa

Hypothesis: What makes a document a "web document" is the ability to reference other documents across the network, either as a source for a component of the document or as a target for the user to retrieve related information.

This can be managed by a small number of attributes:

  • href -- The location of another resource that the element refers to.
  • src -- The location of data that represents this element. If the data is unavailable, the given element contents could be used as an alternate representation.
  • type -- A hint as to the data's MIME type.
  • encoding -- A hint as to the data's text encoding, such as utf-8 or base64.

Any element could theoretically contain these attributes. Therefore, a web document needs no specific elements.

A web document need not even be HTML except for the reason that HTML is all that is currently supported. The XML/SGML superformat is a widely agreed-upon container for web documents, but not necessary except to the extent that it has already become a standard. A future web browser could support multiple web document languages and parse them all to the same general internal tree structure, to which the same styling rules and scripting operations may be applied.

The early years of HTML development were focused on reimplementing features currently present in existing desktop publishing applications and UI component libraries. The current emphasis is on adding semantic tags. At least two alternate XML-based file formats exist that already include a robust set of semantic tags, these being DocBook and OpenDocument. A future standard for web documents could use one of these formats by simply adding to it the web attributes and perhaps also simplifying the standard and loosening up its strictness for the benefit of developers. For example, HTML 6 and Docbook 6 could be the same language.

Discuss.

Page generated Jul. 5th, 2025 08:01 am
Powered by Dreamwidth Studios